When fasted rats ate regular lab chow there was a lag time of about 2 h before the concentration of fructose 2,6-bisphosphate (Fru-2,6-P2) in liver began to rise from its low basal level. By contrast, in animals refed on a sucrose-based diet hepatic [Fru-2,6-P2] increased 20-fold (to a value of approximately 12 nmol/g wet weight) during the first hour. These responses correlated with differences in the ability of the two diets to increase the circulating [insulin]/[glucagon] ratio and thus to elevate the ratio of 6-phosphofructo-2-kinase to fructose-2, 6-bisphosphatase. Liver glycogen was deposited briskly in both groups of rats. To assess its mechanism of synthesis (directly from glucose versus indirectly via the gluconeogenic pathway), animals eating the chow or sucrose diets received intravenous infusions of [14C]bicarbonate, [1-14C] fructose, and 3H2O. After isolation, the glycogen was subjected to positional isotopic analysis of its glucose residues. The results established that regardless of the diet the bulk of liver glycogen was gluconeogenic in origin. The fact that with sucrose feeding carbon flow through hepatic fructose-1,6-bisphosphatase remained active despite high levels of Fru-2,6-P2 (a potent inhibitor of this enzyme in vitro) presents a metabolic paradox. Conceivably, the suppressive effect of Fru-2, 6-P2 on hepatic fructose-1,6-bisphosphatase is overridden in vivo by some unknown factor or factors generated in response to sucrose feeding. Alternatively, metabolic zonation in liver might result in the coexistence of hepatocytes rich in Fru-2,6-P2 (high glycolytic, low gluconeogenic, low glycogenic capacitites) with cells depleted of Fru-2,6-P2 (low glycolytic, high gluconeogenic, high glycogenic capacities).